Health Education Research Advance Access originally published online on September 18, 2006
Health Education Research 2007 22(3):414-424; doi:10.1093/her/cyl097
Out of context? Translating evidence from the North Karelia project over place and time
1 Department of Community Health Sciences
2 Institute of Health Economics and Centre for Health and Policy Studies, University of Calgary, Alberta, Canada T2N 4N1
* Correspondence to: L. McLaren. E-mail: lmclaren{at}ucalgary.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
Within the literature on community-based heart health promotion and chronic disease prevention, the North Karelia project is often viewed as a model program for achieving community-wide reductions in risk factors and mortality associated with cardiovascular disease. In the present study, we examine the tendency to attempt replication of elements of the North Karelia project, without due consideration of the unique population and setting being targeted. We analysed a sample of 64 articles reporting on community-based interventions targeting chronic disease, published between 1990 and 2002. Of these 64 articles, 43 (67%) made explicit reference to North Karelia or one of the other early projects (Stanford, Minnesota, Pawtucket). Of these 43 articles, 8 (19%) explicitly acknowledged the unique features of the population/setting in question, and articulated a need to adapt to these unique features, while 10 (23%) provided no acknowledgment of unique population/setting features. The remaining 25 (58%) were ‘in between’, and examples from each group are discussed. We conclude that for many contemporary community-based interventions, concern with replicating the North Karelia project is accompanied by inadequate consideration or reporting of the details of the unique context (including people, place and time), and this may undermine the success of community-based health promotion.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
Within the literature on community interventions for heart health promotion and chronic disease prevention, the North Karelia project is often viewed as a success [1–5]. This view is supported by long-term statistics on improvements in cause-specific mortality rates over the time period during and following the intervention. In particular, this comprehensive community intervention, which was designed to achieve population-wide changes in dietary habits and ultimately coronary heart disease rates in one province in Finland, has been credited with achieving a 73% reduction in age-adjusted coronary heart disease mortality rate between 1969 and 1995 in this region [6]. The intervention began in 1972 in response to growing public concern about the health profile of the province (heart disease mortality rates for men in the region were the highest in the world). The theoretical framework for the project incorporated behavior change, communication and community organization principles; and the intervention itself included media campaigns, collaboration with the food industry, involvement of local health care and community organizations and agricultural reforms, among other components (for more detail on the North Karelia Project, the reader is referred to [7]). The comprehensive nature of this project, coupled with its direct attention to social, cultural and economic influences on health, was impressive.
Early reports of success from North Karelia inspired many other programs, including three community-based heart health demonstration trials in the United States funded by the National Heart, Lung and Blood Institute: the Stanford Five-City Project, the Minnesota Heart Health Project and the Pawtucket Heart Health Project (for an overview, see [8]). These American projects—though conceived independently—were developed collaboratively with respect to methods, which resulted in a relatively high degree of comparability [9]. All of these projects (including North Karelia) adopted a quasi-experimental design strategy that incorporated one or more intervention communities and one or more (usually matched) reference communities [8].
Partly instigated by disappointing results from these American interventions [9], a sizeable literature consequently emerged that aimed at dissecting these experiences [8, 9]. Though this literature is multifaceted, a prominent feature is debate over whether the controlled trial design is an appropriate evaluation tool for community-based health promotion interventions. As articulated by Mittelmark [9] among others, there are a number of problems that accompany the ‘scaling up’ of an experimental design to the community level. These include the logistical challenges of achieving randomization and control (hallmarks of the controlled trial design) and the weak statistical power to detect intervention effects that stems from a low number of community units and consequently large standard errors of measurement. A suggested alternative to the trial design, for the purpose of community-based interventions, is application of community development approaches [9].
In this paper, we do not take issue with the trial approach per se. Rather, our aim is to draw attention to the tendencies that may accompany use of the controlled trial format. In particular, we are concerned that the desire to adhere to a trial design is accompanied by implicit acceptance of an ‘efficacy’ model of evidence, whereby a particular intervention, once it is confirmed to have ‘worked’ in one context, may be applied in the same format elsewhere. This position is rooted in a rich literature on external validity and related concepts of generalizability, relevance and translation–dissemination–implementation in the bridging of science and practice [10–13], including discussion around the applicability of evidence-based guidelines derived from the evidence-based medicine movement and systematic reviews of literature that put most of their weight on internal validity [14]. As articulated by scholars of the nature of evidence in public health [15, 16], this way of thinking is not appropriate for complex interventions such as community-based health promotion, and there is a need to achieve a better balance between internal validity and external validity in this arena. Contexts vary, and interventions must acknowledge and adapt to the unique circumstances of the population and setting being targeted [17].
Central to this argument is the notion of what constitutes risk. Data suggesting that, for example, increasing body mass index (BMI) is associated with increasing risk for Type 2 diabetes is reasonably straightforward. However, things become more complicated when one considers evidence that risk associated with excess body weight begins at a lower BMI in China, than in the US [18]. Such evidence highlights the fact that risk—even the same risk—cannot be dealt with the same way in different contexts. Furthermore, as articulated by Rose [19], there is typically no clear line between risk that is ‘high’ and that which is ‘moderate’. And for many chronic diseases, there is a much larger number of people at moderate risk than at high risk, and therefore most of the cases (those who develop the illness) will come from those at moderate risk. A moderate lowering of risk across this large number of people thus promises to have a bigger public health impact, than large risk reductions among the few at high risk, and therefore an intervention that aims to achieve community-wide health improvements should pursue moderate risk reduction across much of the population. In order to achieve this, it is essential to target contextual features to which the community as a whole is exposed—i.e. the social, physical, economic, cultural and political aspects of population and its setting.
In principle, two things are needed to achieve an effective intervention: (i) to do the right thing and (ii) to do enough of it. While the latter (i.e. achieving the correct preventive dose) is an important consideration, it is irrelevant if (i) has not been achieved. In the present paper, we are concerned with this first issue: whether interventions are doing the right thing, for the right people, in the right place, at the right time. Our position is that, despite holding community-wide health improvements as an intervention goal, many published intervention studies do not adequately account for, or report on, the contextual (social, physical, economic, political and cultural) features of the population and setting being targeted. In the context of an earlier review of the literature on integrated approaches to the prevention of obesity and chronic disease at the community level, we noticed that it was quite common for authors to cite the North Karelia project as evidence for the success of community intervention studies, as though success does not depend on context (people, place and time). In the present study, we aimed to investigate this tendency. In particular, of studies that reference the North Karelia project, to what extent do authors recognize and report on the potentially modifying role of context? To what extent do authors highlight the unique circumstances of their target population and setting, which may limit the relevance of principles from North Karelia? Can we document variation in the degree to which these features are recognized?
We acknowledge that citing the North Karelia project does not necessarily mean that the author is intending to replicate this older intervention. There may be a tendency to cite North Karelia simply because it is such a prominent landmark in the history of heart health promotion/disease prevention interventions. However, in-line with critics of the predominant focus on internal validity [14], we suggest that citing North Karelia in a relatively uncritical manner nonetheless perpetuates the message that if evidence can show that a community intervention has worked in the past, then that is reason enough to continue to do this today, in a different setting. We feel that to cite North Karelia without exploring and articulating ways in which current circumstances differ is to do a disservice to the community intervention literature by perpetuating the tendency to prioritize internal validity over a deep understanding of local context and culture. Such a narrow model of evidence does not take into account the thoroughly social nature of risks to human health, and is based on a faulty assumption that contextual factors (people, place and time) are not fundamental matters, or that it is possible to ‘control’ for them.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
The sample of articles examined in the present study was originally collected for a review of the literature on integrated approaches to the prevention of obesity and chronic disease (McLaren et al. 2005; available online at http://www.chaps.ucalgary.ca/Working_papers/Report_BRIEF_McLaren.pdf). Articles sought for inclusion in this earlier project were published in English or French between 1990 and 2002, and details of the search strategy used are available from the authors. The present study focuses on integrated health promotion and disease prevention interventions that took place in communities. Please note that from this point onward, ‘health promotion’ is meant to imply both health promotion and disease prevention.
For each community-based study, one published article was selected that appeared to provide the most detail on methods, particularly project design and implementation. Each of these articles was examined for explicit reference to one of the early heart health promotion projects: North Karelia, Stanford, Pawtucket or Minnesota. By explicit reference, we mean mention of these projects by name in the article text or non-specific mention of these projects in the text (e.g. ‘previous community intervention trials’) accompanied by specific project references in the reference list. As mentioned earlier, it is clear from the literature that these projects are ‘the main exemplars from which lessons are drawn about disease prevention practice and research at the community-wide level’ [9, p. 5] in the context of heart health. Studies that did not make reference to any of these projects were excluded from further analysis.
Studies that referred to one or more of these earlier projects were further examined, to assess the extent to which authors recognized and reported on the unique features of the population and setting being targeted. We were interested in whether authors recognized that these unique features would reduce the extent to which principles from earlier trials could be directly transferred. Judgments were made along the two dimensions outlined below. All four authors independently evaluated a pilot sample of five articles, and a consensus judgment was achieved through a team meeting. The remaining articles were evaluated by one author each, and consistency was maximized through discussion with the first author in all cases.
Do the study authors recognize the uniqueness of the population/setting being targeted?
Here, we were interested in the attention paid to characteristics (sociodemographic, economic, cultural and political) of the population/setting being targeted. Judgments included ‘yes, explicitly’ (i.e. the study provided information about the population—such as socioeconomic or ethnic diversity, cultural aspects, etc.—which is recognized for its uniqueness); ‘yes, implicitly’ (same as above, but uniqueness is not explicit) and ‘no’ (little or no information is provided about the population).
Do the study authors recognize the need for adaptation of the intervention to the unique circumstances of the target population/setting?
Here, we assessed whether recognition of the unique features of population/setting was translated into an intervention strategy that was adapted to these unique features. Judgments included yes, explicitly (i.e. the authors explicitly recognized the need to adapt their intervention to the unique population/setting, and articulated their efforts to do so); yes, implicitly (i.e. recognition of the need to adapt is implied, for example, by incorporation of community development principles) and no (no indication of any adaptation).
Based on consensus judgments on the above two dimensions, studies were arranged into three groups. One group included those studies that were judged to have explicitly recognized both the uniqueness of the population/setting being studied and the need for adaptation to these unique circumstances. Another group included those studies for which virtually no information was provided by authors regarding either of these questions. The final group included those studies that were ‘in between’—that is, for which the two questions above had responses of yes implicitly or some combination of yes explicitly, yes implicitly and no.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
A total of 64 articles were examined, and 43 of these (67%) referenced North Karelia, Stanford, Pawtucket or Minnesota. The remaining studies (n = 11, 32%) did not reference any of these studies and were not examined further for the purpose of this paper. Of the 43 studies that referenced one of the early heart health projects, North Karelia was the project cited most often.
Of the 43 studies, eight (19%) made explicit recognition of the unique features of the population/setting in question, and acknowledged the need to adapt to these unique circumstances. Ten (23%) provided no acknowledgment of the unique features of the population/setting in question or did they acknowledge any need to adapt to the unique circumstances. The remaining 25 (58%) were in between. These groupings are provided in Table I, and examples are discussed below.
|
Studies with explicit acknowledgment of unique population/setting and need to adapt
An example of a study in this grouping is the East Harlem Healthy Heart Program [20]. In this article, the following reference to early community interventions was made:
The community intervention model of health promotion and illness prevention is a promising strategy to attack the chronic disease epidemics of industrialised societies in the late 20th century ... First implemented in North Karelia, Finland in the 1970s and replicated in Pawtucket, Rhode Island; Minneapolis, Minnesota and Stanford, California in the 1980s, community intervention is a primary prevention model which seeks to change attitudes, norms, and values regarding behaviors that contribute to chronic disease within a defined population through initiating changes to the social, educational, cultural and physical environment (p. 360).
The author thus situates herself within the existing community intervention literature. However, she departs from this literature by pointing out that this model has ‘not [been] widely tested in urban, low income African American and Latino communities where chronic disease epidemics are virulent’ (p. 360). Unique characteristics of the community of interest are then elaborated; for example:
East Harlem is a predominantly low income, working class community of over 110,000 ... its multi-ethnic population is 53% Latino, predominantly Puerto Rican, and 39% African-American. Spanish language and culture predominate. There are a growing number of undocumented Mexican and Dominican families entering the community as well. Poverty, low education levels, high unemployment, inadequate housing, crime and high rates of substance abuse, AIDS, teen pregnancy and many preventable chronic illnesses are present (p. 362).
Such comments indicate awareness of the importance of the unique circumstances of this population and setting. Furthermore, Brenner [20] explicitly acknowledges the need to take these unique circumstances into account when designing an intervention, as illustrated in the following quotations:
Traditional methods of health education and public health messages in the mass media are usually based on the dominant or mainstream culture and do not reflect the norms, values and/or role models of poor, culturally diverse communities (p. 361).Special sensitivities and applications are required in adapting health promotion technology to ethnic minority communities and special population groups. What worked in Minnesota will not work in Melbourne, Australia,Miami, South Central Los Angeles or East Harlem (p. 362).
In sum, we judged the East Harlem Healthy Heart Program [20] as having incorporated context (uniqueness of people and place) to a relatively high degree. Seven other interventions were judged to fall into this category (see Table I).
Studies with virtually no acknowledgment of unique population/setting or need to adapt
An example of a study in this grouping is Heartbeat Wales [21]. In this article, the following reference to early community interventions was made:
The evaluation of the North Karelia programme was based on comparisons between a single intervention and single reference community ... As a result of the difficulties experienced in the North Karelia program, which shared a common boundary with its chosen reference area, a geographically separate area was deliberately selected to reduce problems of media overspill and more general contamination that might occur along a shared border (p. 128).
This reference to North Karelia is methodological in nature, conveying an assumption that the viability of community intervention programs is dictated by study design features that allow effects to be seen. Though we do not dispute the value of applying principles of experimental design to community interventions, we do take issue with studies for which adherence to study design is prioritized to such an extent that the unique features of population and setting are not discussed and may not have been considered. This approach is illustrated in the Heartbeat Wales study. In the article examined [21], there is no information provided about the sociodemographic, political, cultural, or economic features of the population or setting. In fact, the only comment about the population is: ‘[relative to communities targeted in previous intervention programs] Wales is a comparatively large country ... with a substantially greater population’ (p. 128). There is no mention of any need to tailor or adapt the intervention to the local environment or social geography.
In sum, we judged Heartbeat Wales [21] as not having incorporated context to any discernable degree. Since no information was provided about the population and setting in which it was implemented, this study appeared to have reproduced a generic or standard intervention program. We were led to infer that the authors sought to ensure the presence of study design features that would improve the likelihood of detecting an effect, as opposed to improving the likelihood of having an effect. Nine other interventions were judged to fall into this category (see Table I).
Studies with limited acknowledgment of unique population/setting or need to adapt (in-between category)
Because the in-between category was more diverse than those above, two examples will be provided. The first is Keeping the Heart Beat in Grampian [22]. This article made the following reference to early studies:
The Grampian Heart campaign, a 10 year initiative, is partly based on many of the theoretical principles which have been tried successfully in the USA, Finland, and Wales. These programs, although community based in their orientation, have stopped short of handing over any real responsibility to the community. They have been in this sense professionally led and controlled (p. 13).
The authors thus introduce their study by drawing attention to a perceived deficit in earlier studies: insufficient attention to community ownership. While this would suggest attention to the unique characteristics of the people and place being targeted, in fact no information was provided about the population or setting in Grampian.
Through an emphasis on community ownership principles, Macallan and Narayan [22] implicitly recognize the need to adapt to local context. This is apparent from the following quotation:
The Grampian Heart Campaign ... is unique and innovative in a number of important respects. It has been registered as a company limited by guarantee with charitable status to give it a corporate identity and so that ownership, responsibility and control are truly vested in the community (pp. 13, 14).
Although community ownership is prioritized, we are concerned about the absence of information provided about the local population and setting. The absence of such information makes it difficult for readers to judge whether it might be worthwhile to use similar techniques to achieve community ownership in another context. Therefore, although community involvement provides a means for program planners to understand the population and adapt the program accordingly, we are concerned about the lack of information reported on context, and thus judged that context is not really accounted for in this intervention.
A second example from this in-between category is the Coronary Risk Factor Study (CORIS) [23]. These authors made the following reference to early studies:
To date only three adequately-evaluated community programmes have published their final results, namely the Stanford Three-Community Study, the North Karelia Project, and the Stanford Five-City Project. The results of these studies were encouraging, in that reductions in overall risk were achieved (p. 428).
This comment draws attention to the fact that Rossouw et al. [23] view these earlier programs as at least somewhat successful. Our pattern of judgments about this study is different from comments about the Grampian study above. In particular, authors of the CORIS study provided some information about the population being targeted for intervention; in particular, that ‘the adult populations were largely middle-class and farming and farming-related service activities were the main occupations’ (p. 429). On the other hand, there was no recognition of the need to tailor or adapt the intervention to this unique Afrikaans-speaking population of South Africa. In the discussion section, it is acknowledged that ‘it is possible that the behavioural norm of the community sets the limit to the degree of change that can be achieved in a set time period’ (p. 437), but this is a post hoc explanation for small magnitude of effects.
In sum, Keeping the Heart Beat in Grampian and CORIS were judged to be in between with respect to accounting for context. The Grampian study prioritized community ownership principles but provided no information about the Grampian population or setting, whereas the CORIS study described the population briefly but did not follow through on the implications of these unique characteristics for adaptation. Twenty-three other studies were judged to be in this in-between category (see Table I).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
This study aimed to test our impression, formed through the process of carrying out a comprehensive literature review, that much of the literature on community-based disease prevention interventions endorses a model of evidence whereby replication of earlier trials is prioritized over attention to unique features of the current population and setting. Of the articles in our sample that referenced North Karelia or another early health promotion trial, we found that only 19% of studies departed notably from this approach. In other words, fewer than one in five studies in our sample devoted significant space to reporting details about the unique features of people, place and time; even though these features are considered in various literatures to be essential to understanding and enabling health [24, 25]. Thus, it seems appropriate to conclude that, within the literature sampled (which included studies that adopted a range of strategies and frameworks), the norm is to emphasize certain study design elements—in particular, adherence to a trial design and associated pursuit of internal validity—over contextual features and a strong understanding of local culture.
This underemphasis of context is somewhat surprising, given the explicit incorporation of context in health promotion models [26] and the renewal of interest in an ecological approach to health, which inherently incorporates contextual factors [27]. Further, it is widely accepted that unique features of person, place and time played a large role in the outcomes of the early community heart health trials. The context of the North Karelia project, for example, was such that heart disease rates were high and stable. Notwithstanding various obstacles to the intervention (e.g. socioeconomic deprivation, limited health resources, social norms around diet), one advantage of the historical context was that there was simply more room for change, relative to a contemporary intervention that would be working against a secular decline in heart disease rates [28, 29]. Along these lines, conclusions of a review by Schooler et al. [8] indicate that the effects of interventions that began in the 1970s had more consistent effects than those conducted in the 1980s, and it is suggested that this had to do with the increases in general awareness of risk factors for cardiovascular disease. Thus, the evolution of cardiovascular disease, including increasing public awareness of risk factors and secular trends in prevalence over the past several decades, indicates a context that was very different for North Karelia than for subsequent and contemporary interventions.
While the failure of subsequent trials including Stanford, Minnesota and Pawtucket has been acknowledged [30, 31] (though not universally; e.g. [8, 32]), the predominant interpretation carries an assumption that the intervention itself is effective and that the lack of impact reflects factors such as insufficient duration, intensity of intervention delivery or uptake, inadequate statistical power to detect an effect or insensitive evaluation tools [29, 33]. While these aspects may be playing some role, we hope to have conveyed that lack of attention to context is a very credible explanation for the underwhelming success of community-based heart health promotion. In-line with others [16, 34], we feel that the predominant approach demonstrated in our work needs to be rethought to ensure that explicit attention is paid to contextual variation from one intervention site to another.
For this research, we relied on one published article to represent each intervention project. Though we attempted to select the article with the most detail about study design, the possibility remains that the contextual material of most interest to us was available elsewhere or was pre-empted due to constraints on space or style imposed by many peer-reviewed journals. Although our reliance on one article may have influenced our results, we do not believe that this possibility represents a major threat to the validity of our findings, since regardless of constraints in journals, authors must make a judgment as to what material is important enough to include. Even if there is insufficient space to detail the unique context of the intervention, authors might refer the reader to other published articles or online documentation, and we trust that authors who are concerned about these issues will do so. Insufficient reporting of contextual details makes it difficult or impossible for a reader to judge the transferability of research to a new context. However, we are sympathetic to the norms of publishing in health and medical journals, for which there is often a disincentive to devote adequate attention to local culture and context, since such work is not understood to be generalizable in a methodological or statistical sense. Thus, we advocate for greater acceptance of such material on the part of peer review committees, and fortunately, there appears to be progress in this direction. Discussion of criteria and guidelines to strike a better balance between internal and external validity is evident in both peer-reviewed articles [14] and formal interactions among journal editors http://obssr.od.nih.gov/Content/Conferences_And_Workshops/Conference_FY2004/Complex_Interventions.htm).
In conclusion, there is a notable tendency in the community-based health promotion literature to reference early heart health trials, and to attempt replication of these early trials without due consideration of the unique people, place and time being targeted. Notwithstanding the opposite problem of the potential for overadapting, as discussed by Green and Kreuter [35], our contribution in this work is to draw attention to the practice of underadapting due to dearth of attention to context as illustrated in the majority of studies we reviewed. Given the somewhat underwhelming evidence for the success of community-based health promotion interventions [30, 31, 36], we concur with others [14] on the need for an approach to the design and reporting of these interventions that places much more emphasis on the importance of context than presently observed.
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
None declared.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
This work is based on an earlier project entitled ‘Are integrated approaches working to promote healthy weights and prevent obesity and chronic disease?’, for which financial contribution was provided by Health Canada's Health Policy Research Program. Authors would like to acknowledge the following sources of funding: Alberta Heritage Foundation for Medical Research (LM), Canadian Institutes of Health Research (LM) and Markin Health and Society Program (LMG). We would like to thank three anonymous reviewers whose helpful comments contributed to a greatly improved manuscript.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConflict of interest statementAcknowledgementsReferences |
|---|
1. Blashko CA, Paterson JG. You and your heart. Promoting preventive health education through the mass media. Int J Adv Couns (1990) 13:49–59.[CrossRef][ISI]
2. Brownson RC, Mayer JP, Dusseault PM, et al. Developing and evaluating a cardiovascular risk reduction project. Am J Health Behav (1997) 21:333–45.[ISI]
3. Burgess Record N, Harris DE, Record SS, et al. Mortality impact of an integrated community cardiovascular health program. Am J Prev Med (2000) 19:30–8.[ISI][Medline]
4. Pratt CA, Hurst R, Williams KP, et al. Evaluating cardiovascular disease prevention programs in African American communities. J Public Health Manag Pract (1999) 5:81–90.[Medline]
5. Shea S, Basch CE, Lantigua R, et al. The Washington Heights-Inwood Healthy Heart Program: a third generation community-based cardiovascular disease prevention program in a disadvantaged urban setting. Prev Med (1992) 21:203–17.[CrossRef][ISI][Medline]
6. Puska P. The North Karelia Project: from community intervention to national activity in lowering cholesterol levels and CHD risk. Eur Heart J (1999) 1(Suppl.):S9–13.
7. Puska P, Tuomilehto J, Nissinen A, et al, eds. The North Karelia Project. 20 Year Results and Experiences (1995) Helsinki: Finland University Press.
8. Schooler C, Farquhar JW, Fortmann SP, et al. Synthesis of findings and issues from community prevention trials. Ann Epidemiol (1997) S7:S54–68.[CrossRef][ISI]
9. Mittelmark MB. Health promotion at the communitywide level. In: Health Promotion at the Community Level: New Advances—Bracht N, ed. (1999) Thousand Oaks, CA: Sage. 3–27.
10. Campbell DT, Stanley JC. Experimental and Quasi-Experimental Designs for Research on Teaching (1963) Chicago, IL: Rand McNally.
11. Cronbach LH, Glesser GC, Nanda H, et al. The Dependability of Behavioral Measurements: Theory of Generalizability for Scores and Profiles (1972) New York: John Wiley.
12. Green LW. From research to "best practices" in other settings and populations. Am J Health Behav (2001) 25:165–78.[ISI][Medline]
13. Shadish WR, Cook TD, Campbell DT. Experimental and Quasi-Experimental Design for Generalized Causal Inference (2002) Boston, MA: Houghton Mifflin.
14. Green LW, Glasgow RE. Evaluating the relevance, generalization, and applicability of research: issues in external validation and translation methodology. Eval Health Prof (2006) 29:126–53.
15. Petticrew M, Roberts H. Evidence, hierarchies, and typologies: horses for courses. J Epidemiol Community Health (2003) 57:527–29.
16. Rychetnik L, Frommer M, Hawe P, et al. Criteria for evaluating evidence on public health interventions. J Epidemiol Community Health (2002) 56:119–27.
17. Hawe P, Shiell A, Riley T. Complex interventions: how "out of control" can a randomised controlled trial be? Br Med J (2004) 328:1561–3.
18. Wildman RP, Gu D, Reynolds K, et al. Appropriate body mass index and waist circumference cutoffs for categorization of overweight and central adiposity among Chinese adults. Am J Clin Nutr (2004) 80:1129–36.
19. Rose G. The Strategy of Preventive Medicine (1992) Oxford: Oxford University Press.
20. Brenner B. Engaging in a public health program with populations that are the traditional targets of social work intervention. Soc Work Health Care (2002) 35:359–75.[CrossRef][ISI][Medline]
21. Nutbeam D, Smith C, Murphy S, et al. Maintaining evaluation designs in long term community based health promotion programmes: Heartbeat Wales case study. J Epidemiol Community Health (1993) 47:127–33.[Abstract]
22. Macallan L, Narayan V. Keeping the heart beat in Grampian—a case study in community participation and ownership. Health Promot Int (1994) 9:13–9.
23. Rossouw JE, Jooste PL, Chalton DO, et al. Community-based intervention: the Coronary Risk Factor Study (CORIS). Int J Epidemiol (1993) 22:428–38.
24. Evans RG, Barer ML, Marmor TR, eds. Why are Some People Healthy and Others Not? The Determinants of Health of Populations (1994) New York: Aldine de Gruyter.
25. Minkler M. Health education, health promotion and the open society: an historical perspective. Health Educ Q (1989) 16:17–30.[ISI][Medline]
26. Ottawa Charter for Health Promotion (1986) WHO/HPR/HEP/95.1. Geneva: World Health Organisation.
27. McLaren L, Hawe P. Ecological perspectives in health research: a glossary. J Epidemiol Community Health (2005) 59:6–14.
28. Green LW. Community health promotion: applying the science of evaluation to the initial sprint of a marathon. Am J Prev Med (1997) 13:225–8.[ISI][Medline]
29. Susser M. Editorial: the tribulations of trials—interventions in communities. Am J Public Health (1995) 85:156–8.
30. Ebrahim S, Davey Smith G. Exporting failure? Coronary heart disease and stroke in developing countries. Int J Epidemiol (2001) 30:201–5.
31. Winkleby MA, Feldman HA, Murray DM. Joint analysis of three U.S. community intervention trials for reduction of cardiovascular disease risk. J Clin Epidemiol (1997) 50:645–58.[CrossRef][ISI][Medline]
32. Shea S, Basch CD. A review of five major community-based cardiovascular disease prevention programs. Part II: Intervention strategies, evaluation methods, and results. Am J Health Promot (1990) 4:279–87.[Medline]
33. Sellers DE, Crawford SL, Bullock K, et al. Understanding the variability in the effectiveness of community heart health programs: a meta-analysis. Soc Sci Med (1997) 44:1325–39.[CrossRef][ISI][Medline]
34. Potvin L, Gendron S, Bilodeau A, et al. Integrating social theory into public health practice. Am J Public Health (2005) 95:591–5.
35. Green LW, Kreuter MW. Fighting back or fighting themselves? Community coalitions against substance abuse and their use of best practices. Am J Prev Med (2002) 23:303–6.[CrossRef][ISI][Medline]
36. Syme SL. Social determinants of health: the community as an empowered partner. Prev Chronic Dis (2004) 1:1–5.
37. Higginbotham N, Heading G, McElduff P, et al. Reducing coronary heart disease in the Australian Coalfields: evaluation of a 10-year community intervention. Soc Sci Med (1999) 48:683–92.[CrossRef][ISI][Medline]
38. O'Loughlin J, Paradis G, Kishchuk N, et al. Coeur en sante St-Henri—a heart health promotion programme in Montreal, Canada: design and methods for evaluation. J Epidemiol Community Health (1995) 49:495–502.[Abstract]
39. Renaud L, O'Loughlin J, Paradis G, et al. Un programme de promotion de la sante cardiovasculaire aupres des 9–12 ans et de la communaute Saint-Louis du Parc/Quebec. Sante Publique (1998) 10:425–45.[Medline]
40. Daniel M, Green LW, Marion SA, et al. Effectiveness of community-directed diabetes prevention and control in a rural Aboriginal population in British Columbia, Canada. Soc Sci Med (1999) 48:815–32.[CrossRef][ISI][Medline]
41. Nafziger AN, Erb TA, Jenkins PL, et al. The Otsego-Schoharie healthy heart program: prevention of cardiovascular disease in the rural US. Scand J Public Health (2001) 29:21–32.
42. Hood VL, Kelly B. A native American community initiative to prevent diabetes. Ethnicity Health (1997) 2:277–85.[Medline]
43. Baxter T, Milner P, Wilson K, et al. A cost effective, community based heart health promotion project in England: prospective comparative study. Br Med J (1997) 315:582–5.
44. Goodman RM, Wheeler FC, Lee PR. Evaluation of the Heart to Heart Project: lessons from a community-based chronic disease prevention project. Am J Health Promot (1995) 9:443–55.[ISI][Medline]
45. Andersson CM, Bjaras GEM, Ostenson CG. A stage model for assessing a community-based diabetes prevention program in Sweden. Health Promot Int (2002) 17:317–27.
46. Wiesemann A, Metz J, Nuessel E, et al. Four years of practice-based and exercise-supported behavioural medicine in one community of the German CINDI area. Int J Sports Med (1997) 18:308–15.[ISI][Medline]
47. Osler M, Jespersen NB. The effect of a community-based cardiovascular disease prevention project in a Danish municipality. Dan Med Bull (1993) 40:485–9.[ISI][Medline]
48. Foerster SB, Kizer KW, DiSogra LK, et al. California's "5 a day—for better health" campaign: an innovative population-based effort to effect large-scale dietary change. Am J Prev Med (1995) 11:124–31.[ISI][Medline]
49. Hoffmeister H, Mensink GBM, Stolzenberg H, et al. Reduction of coronary heart disease risk factors in the German cardiovascular prevention study. Prev Med (1996) 25:135–45.[CrossRef][ISI][Medline]
50. Dunbar JA, Vincent DS, Dunbar AP, et al. Implementing the North Karelia project in Scotland. Scott Med J (1998) 43:99–101.[ISI][Medline]
51. Muntoni S, Stabilini L, Stabilini M, et al. Results of a five-year community-based programme for cardiovascular disease prevention: the ATS-Sardegna campaign. Eur J Epidemiol (1999) 15:29–34.[CrossRef][ISI][Medline]
52. Naylor PJ, Wharf-Higins J, Blair L, et al. Evaluating the participatory process in a community-based heart health project. Soc Sci Med (2002) 55:1173–87.[CrossRef][ISI][Medline]
53. Potter JD, Graves KL, Finnegan JR, et al. The Cancer and Diet Intervention Project: a community-based intervention to reduce nutrition-related risk of cancer. Health Educ Res (1990) 5:489–503.
54. Angelico F, Del Ben M, Giampaoli S, et al. The "Di.S.Co." project: a community-based comprehensive control project of chronic diseases. Time change in intervention strategy. Ann Ig (1997) 9:101–10.[Medline]
55. Nader PR, Sallis JF, Abramson IS, et al. Family-based cardiovascular risk reduction education among Mexican- and Anglo-Americans. Fam Community Health (1992) 15:57–74.
56. Bental T, Paran E. Increasing public awareness of the risk of cardiovascular disease using ‘health activists’ in a community health project in Israel: a feasibility study. J Hum Hypertens (1990) 4:356–61.[ISI][Medline]
57. Van Assema P, Steenbakkers M, Kok G, et al. Results of the Dutch community project "Healthy Bergeyk". Prev Med (1994) 23:394–401.[CrossRef][ISI][Medline]
58. Komarek L, Kebza V, Lhotska L, et al. "Healthy Dubec"—design of a joint Czech–American community project for the reduction of cardiovascular and cerebrovascular disease (adapted from American experience). Cent Eur J Public Health (1995) 4:230–3.
59. Fonnebo V, Sogaard AJ. The penetrating educational effect of a mass-media based fundraising campaign "Heart for Life". Scand J Soc Med (1990) 18:185–93.[ISI][Medline]
60. Isacsson A, Lindholm LH, Schersten B, et al. Community intervention against non-insulin-dependent diabetes mellitus (NIDDM) and cardiovascular disease: a study based on Swedish health care. Cardiovasc Risk Factors (1996) 6:164–71.
61. Muller MJ, Asbeck I, Mast M, et al. Prevention of obesity—more than an intention. Concept and first results of the Kiel Obesity Prevention Study (KOPS). Int J Obes (2001) 25(Suppl. 1):S66–74.[CrossRef][ISI]
62. Shelley E, Daly L, Graham A, et al. The Kilkenny Health Project: a community research and demonstration cardiovascular health programme. Ir J Med Sci (1991) 60(Suppl. 9):10–6.
63. Sommerlad EA, Robbins A. Healthy alliances and social action broadcasting: assessment of a local radio project. Health Educ J (1997) 56:51–63.
64. Metcalf CA, Hoffman MN, Steyn K, et al. Design and baseline characteristics of a hypertension intervention program in a South African village. J Hum Hypertens (1996) 10:21–6.[ISI][Medline]
65. Dowse GK, Gareeboo H, Alberti KGMM, et al. Changes in population cholesterol concentrations and other cardiovascular risk factor levels after five years of the non-communicable disease intervention programme in Mauritius. Br Med J (1995) 311:1255–9.
66. Cutter J, Tan BY, Suok KC. Levels of cardiovascular disease risk factors in Singapore following a national intervention programme. Bull World Health Organ (2001) 79:908–15.[ISI][Medline]
67. Weinehall L, Hellsten G, Boman K, et al. Prevention of cardiovascular disease in Sweden: the Norsjo community intervention programme—motives, methods and intervention components. Scand J Public Health (2001) 29(Suppl. 56):13–20.[ISI][Medline]
68. Lupton BS, Fonnebo V, Sogaard AJ, et al. The Finnmark Intervention Study. Better health for the fishery population in an arctic village in North Norway. Scand J Prim Health Care (2002) 20:213–8.[CrossRef][ISI][Medline]
69. Engelgau MM, Venkat Narayan KM, Geiss LS, et al. A project to reduce the burden of diabetes in the African–American community: Project Direct. J Natl Med Assoc (1998) 90:605–13.[Medline]
70. Pelletier J, Moisan J, Roussel R, et al. Heart health promotion: a community development experiment in a rural area of Quebec, Canada. Health Promot Int (1997) 12:291–8.
Received on March 22, 2006; accepted on August 11, 2006
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||